GENETICA: A computer language that supports general formal expression with evolving data structures

This paper presents a general problem-solving method combining the principles of artificial intelligence and evolutionary computation. The problem-solving method is based on the computer language GENETICA, which stands for "Genetic Evolution of Novel Entities Through the Interpretation of Composite Abstractions." GENETICAs programming environment includes a computational system that evolves data abstractions, viewed as genotypes of data generation scenarios for a GENETICA program, with respect to either confirmation or optimization goals. A problem can be formulated as a GENETICA program, while the solution is represented as a data structure resulting from an evolved data generation scenario. This approach to problem solving offers: 1) generality, since it concerns virtually any problem stated in formal logic; 2) effectiveness, since formally expressed problem-solving knowledge can be incorporated in the problem statement; and 3) creativity, since unpredictable solutions can be obtained by evolved data structures. It is shown that domain specific languages, including genetic programming ones, that inherit GENETICAs features can be developed in GENETICA. The language G-CAD, specialized to problem solving in the domain of architectural design, is presented as a case study followed by experimental results.     

Evolutionary design of 2-dimensional recursive filters via the computer language GENETICA

In this paper, we present a new design method for a class of two-dimensional (2-D) recursive digital filters using an evolutionary computational system. The design of the 2-D filter is reduced to a constrained minimization problem the solution of which is achieved by the convergence of an appropriate evolutionary algorithm. In our approach, the genotypes of potential solutions have a uniform probability within the region of the search space specified by the constraints and zero probability outside this region. This approach is particularly effective as the evolutionary search considers only those potential solutions that respect the constraints. We use the computer language GENETICA, which provides the expressive power necessary to get an accurate problem formulation and supports an adjustable evolutionary computational system. Results of this procedure are illustrated by a numerical example, and compared with those of some previous designs.